I
nd
o
ne
s
ia
n J
o
urna
l o
f
E
lect
rica
l En
g
ineering
a
nd
Co
m
pu
t
er
Science
Vo
l.
42
,
No
.
1
,
A
p
r
il
20
26
,
p
p
.
1
42
~1
4
8
I
SS
N:
2
502
-
4
7
52
,
DOI
: 1
0
.
1
1
5
9
1
/ijee
cs
.v
42.
i
1
.
p
p
1
42
-
14
8
142
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ij
ee
cs
.
ia
esco
r
e.
co
m
La
ng
ua
g
e mo
dels
and deep
neural
networks
for Ara
bic nam
ed
entit
y
r
ecog
nition
So
m
ia
K
hedim
i
1,
2
,
Abdelg
ha
ni B
o
uzia
ne
1,
2
1
D
e
p
a
r
t
me
n
t
o
f
C
o
m
p
u
t
e
r
S
c
i
e
n
c
e
,
I
n
st
i
t
u
t
e
o
f
S
c
i
e
n
c
e
,
U
n
i
v
e
r
si
t
y
o
f
N
a
a
m
a
,
N
a
a
ma
,
A
l
g
e
r
i
a
2
EED
I
S
La
b
o
r
a
t
o
r
y
,
D
e
p
a
r
t
me
n
t
o
f
C
o
mp
u
t
e
r
S
c
i
e
n
c
e
,
U
n
i
v
e
r
si
t
y
o
f
D
j
i
l
l
a
l
i
L
i
a
b
e
s,
S
i
d
i
B
e
l
A
b
b
è
s,
A
l
g
e
r
i
a
Art
icle
I
nfo
AB
S
T
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
J
u
l
8
,
2
0
2
5
R
ev
is
ed
Feb
20
,
2
0
2
6
Acc
ep
ted
Ma
r
4
,
2
0
2
6
To
k
e
n
t
y
p
e
id
e
n
ti
fica
ti
o
n
li
e
s
a
t
th
e
c
o
re
o
f
n
a
m
e
d
e
n
ti
t
y
re
c
o
g
n
i
ti
o
n
,
a
ll
o
win
g
m
o
d
e
ls
t
o
d
is
ti
n
g
u
is
h
n
a
m
e
d
e
n
ti
ti
e
s
fr
o
m
n
o
n
-
e
n
ti
t
y
t
o
k
e
n
s
a
n
d
th
e
re
b
y
b
e
tt
e
r
c
a
p
tu
re
se
n
te
n
c
e
m
e
a
n
in
g
.
Th
is
p
a
p
e
r
p
re
se
n
ts
a
d
e
e
p
lea
rn
in
g
a
p
p
ro
a
c
h
f
o
r
th
e
Ara
b
ic
n
a
m
e
d
e
n
ti
t
y
re
c
o
g
n
it
i
o
n
tas
k
,
lev
e
ra
g
in
g
d
e
e
p
n
e
u
ra
l
n
e
two
r
k
s
a
n
d
p
re
train
e
d
l
a
n
g
u
a
g
e
m
o
d
e
ls.
T
h
e
p
ro
p
o
se
d
m
o
d
e
l
is
a
c
o
m
b
in
a
ti
o
n
o
f
t
h
e
Ara
EL
ECT
RA
lan
g
u
a
g
e
m
o
d
e
l
with
t
h
e
b
i
d
irec
ti
o
n
a
l
lo
n
g
sh
o
r
t
-
ter
m
m
e
m
o
ry
(BiL
S
TM
)
n
e
u
ra
l
n
e
two
r
k
.
We
u
ti
li
z
e
th
e
Wo
jo
o
d
NER
d
a
tas
e
t,
w
h
ich
p
ro
v
i
d
e
s
fin
e
-
g
ra
i
n
e
d
a
n
n
o
tatio
n
s
o
f
A
ra
b
ic
tex
t
a
c
ro
ss
2
1
e
n
ti
ty
t
y
p
e
s.
T
h
e
re
su
lt
s
o
f
th
is
a
p
p
r
o
a
c
h
a
re
e
n
c
o
u
ra
g
in
g
,
wit
h
a
n
a
c
c
u
ra
c
y
o
f
9
8
.
2
9
%
a
n
d
a
n
F
1
-
sc
o
re
o
f
8
7
%
.
K
ey
w
o
r
d
s
:
Ar
aE
L
E
C
T
R
A
B
iLST
M
Dee
p
lear
n
in
g
L
an
g
u
ag
e
m
o
d
els
Nam
ed
en
tity
r
ec
o
g
n
itio
n
i
n
Ar
ab
ic
W
o
jo
o
d
d
ataset
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
So
m
ia
Kh
ed
im
i
Dep
ar
tm
en
t o
f
C
o
m
p
u
ter
Scie
n
ce
,
I
n
s
titu
te
o
f
Scien
ce
,
Un
iv
er
s
ity
o
f
Naa
m
a
Naa
m
a,
Alg
er
ia
E
m
ail:
k
h
ed
im
i@
cu
n
iv
-
n
aa
m
a
.
d
z
1.
I
NT
RO
D
UCT
I
O
N
Nam
ed
en
tity
r
ec
o
g
n
itio
n
(
NE
R
)
is
a
s
u
b
task
o
f
n
atu
r
al
la
n
g
u
ag
e
p
r
o
ce
s
s
in
g
(
NL
P)
th
at
f
o
cu
s
es
o
n
id
en
tify
in
g
a
n
d
class
if
y
in
g
n
am
ed
en
titi
es
with
in
a
tex
t
in
to
p
r
e
d
ef
in
ed
ca
te
g
o
r
ie
s
s
u
ch
as
p
eo
p
le,
o
r
g
an
izatio
n
s
,
lo
ca
tio
n
s
,
d
ates,
an
d
o
th
er
s
p
ec
if
ic
tag
s
.
NE
R
is
cr
u
cial
in
ex
tr
ac
tin
g
m
ea
n
in
g
f
u
l
in
f
o
r
m
atio
n
f
r
o
m
u
n
s
tr
u
ctu
r
e
d
tex
t t
h
at
ex
i
s
ts
o
n
th
e
web
an
d
s
o
cial
m
e
d
i
a.
NE
R
i
s
a
f
u
n
d
am
en
tal
task
in
s
ev
er
al
NL
P
ap
p
licatio
n
s
.
Fo
r
ex
am
p
le,
in
teg
r
atin
g
NE
R
in
t
o
m
ac
h
in
e
tr
an
s
latio
n
s
y
s
tem
s
h
elp
s
im
p
r
o
v
e
tr
a
n
s
latio
n
q
u
ality
b
y
ac
c
u
r
ately
id
e
n
tify
in
g
an
d
p
r
eser
v
in
g
n
am
ed
en
titi
es
s
u
ch
as
p
eo
p
le,
p
lace
s
,
an
d
o
r
g
an
izatio
n
s
,
wh
ich
ar
e
o
f
ten
m
is
tr
an
s
lated
o
r
o
m
itted
in
tr
ad
itio
n
al
tr
an
s
latio
n
m
o
d
els
[
1
]
.
NE
R
p
lay
s
a
cr
u
c
ial
r
o
le
in
q
u
esti
o
n
-
an
s
wer
in
g
s
y
s
tem
s
,
en
ab
lin
g
th
e
s
y
s
tem
to
r
etr
iev
e
r
elev
an
t
p
ass
ag
es
to
a
q
u
esti
o
n
b
y
id
en
tify
in
g
th
e
n
am
ed
en
titi
es
in
th
e
q
u
esti
o
n
[
2
]
.
I
n
tex
t
s
u
m
m
ar
izatio
n
,
NE
R
h
elp
s
in
id
en
tify
i
n
g
d
if
f
er
en
t
n
am
ed
en
titi
es to
k
ee
p
t
h
em
in
t
h
e
s
u
m
m
ar
y
[
3
]
.
A
s
ig
n
if
ican
t
p
o
r
tio
n
o
f
NE
R
r
esear
ch
is
d
ed
icate
d
to
E
n
g
lis
h
,
wh
er
ea
s
b
u
ild
in
g
a
r
o
b
u
s
t
Ar
ab
ic
NE
R
s
y
s
tem
s
ti
ll
p
r
esen
ts
g
r
ea
ter
ch
allen
g
es
th
an
in
o
t
h
er
l
an
g
u
ag
es
[
4
]
.
T
h
is
r
ef
er
s
to
s
e
v
er
al
ch
allen
g
es
f
o
r
th
e
Ar
ab
ic
lan
g
u
ag
e
.
Ar
ab
ic
h
as
a
r
ich
m
o
r
p
h
o
lo
g
y
wh
er
e
s
u
f
f
ix
es
an
d
p
r
ef
i
x
es
ca
n
b
e
ad
d
ed
to
wo
r
d
s
,
m
ak
in
g
a
s
in
g
le
wo
r
d
eq
u
iv
a
len
t
to
an
en
tire
E
n
g
lis
h
s
en
ten
ce
[
5
]
.
Fo
r
ex
a
m
p
le,
"
ُ
ه
َ
ن
و
ب
ُ
ت
ْ
ك
َ
ي
َ
س
"
tr
an
s
lates
to
“th
ey
will
wr
ite
it”.
Fu
r
th
e
r
m
o
r
e,
in
Ar
ab
ic,
th
er
e
is
n
o
ca
p
italizatio
n
,
w
h
ich
p
la
y
s
a
cr
u
cial
r
o
le
in
id
en
tify
in
g
p
r
o
p
er
n
o
u
n
s
[
6
]
.
T
o
ad
d
r
ess
th
ese
ch
allen
g
es,
we
p
r
o
p
o
s
e
a
d
ee
p
lear
n
in
g
ap
p
r
o
ac
h
f
o
r
Ar
a
b
ic
n
am
ed
r
e
co
g
n
itio
n
(
ANE
R
)
.
I
n
th
is
ap
p
r
o
ac
h
,
we
em
p
lo
y
th
e
Ar
aE
L
E
C
T
R
A
p
r
etr
ain
ed
lan
g
u
a
g
e
m
o
d
e
l
to
r
ep
r
esen
t
th
e
m
ea
n
in
g
o
f
th
e
Ar
ab
ic
tex
t
as
v
ec
to
r
s
.
T
o
en
r
ic
h
th
is
r
ep
r
es
en
tatio
n
,
we
u
s
e
a
v
ar
ian
t
o
f
th
e
r
ec
u
r
r
e
n
t
n
eu
r
al
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
La
n
g
u
a
g
e
mo
d
els a
n
d
d
ee
p
n
e
u
r
a
l n
etw
o
r
ks fo
r
A
r
a
b
ic
n
a
med
en
tity reco
g
n
itio
n
(
S
o
mia
K
h
ed
imi
)
143
n
etwo
r
k
(
R
NN)
,
wh
ich
is
th
e
b
id
ir
ec
tio
n
al
lo
n
g
s
h
o
r
t
-
ter
m
m
em
o
r
y
(
B
iLST
M)
,
in
v
esti
g
atin
g
th
e
im
p
ac
t
o
f
th
e
C
NN
-
B
iLST
M
c
o
m
b
in
atio
n
o
n
th
e
NE
R
task
.
W
e
o
r
g
an
ize
th
e
r
est
o
f
th
is
p
ap
er
as
f
o
llo
ws:
in
th
e
n
e
x
t
Sectio
n
,
we
r
ev
iew
th
e
ex
is
tin
g
Ar
ab
ic
d
atasets
f
o
r
th
e
ANE
R
task
,
as
well
as
th
e
p
r
o
p
o
s
ed
ap
p
r
o
a
ch
o
f
o
th
er
r
esear
c
h
er
s
.
I
n
Se
ctio
n
3
,
we
d
escr
ib
e
th
e
p
r
o
p
o
s
ed
m
o
d
el
a
n
d
th
e
d
ataset
u
s
ed
.
T
h
e
o
b
tain
ed
r
e
s
u
lts
ar
e
p
r
esen
ted
an
d
d
is
cu
s
s
ed
in
Sectio
n
4
.
Fin
ally
,
we
co
n
clu
d
e
with
a
co
n
clu
s
io
n
an
d
f
u
tu
r
e
p
er
s
p
ec
tiv
es.
2.
RE
L
AT
E
D
WO
RK
Sev
er
al
d
atasets
an
d
m
o
d
els
wer
e
cr
ea
ted
f
o
r
th
e
Ar
a
b
ic
NE
R
task
.
I
n
th
is
s
ec
tio
n
,
we
r
ev
iew
th
e
m
o
s
t c
o
m
m
o
n
l
y
u
s
ed
d
atasets
as we
ll a
s
th
e
m
o
s
t r
ec
en
t m
o
d
els.
2
.
1
.
Ara
bic
na
m
ed
ent
it
y
da
t
a
s
et
s
B
en
ajib
a
et
a
l
.
[
7
]
co
llected
3
1
6
ar
ticles
f
r
o
m
Ar
ab
ic
n
e
wsp
ap
er
s
to
cr
ea
te
ANE
R
C
O
R
P
,
wh
ich
co
n
tain
s
1
5
0
,
2
8
6
t
o
k
en
s
ca
te
g
o
r
ized
in
t
o
f
o
u
r
en
titi
es:
lo
ca
tio
n
(
L
OC
)
,
p
er
s
o
n
(
PERS
)
,
o
r
g
an
is
atio
n
(
OR
G)
,
an
d
Miscellan
eo
u
s
(
MI
SC
)
.
C
L
E
ANAN
E
R
C
o
r
p
:
is
a
clea
n
e
d
v
er
s
io
n
o
f
th
e
ANE
R
C
O
R
P
d
ataset
r
elea
s
ed
b
y
Al
-
Du
wais
et
a
l
.
[
8
]
.
T
h
ey
ad
d
1
.
3
3
% m
is
s
in
g
lab
els an
d
co
r
r
ec
t 5
.
1
1
% in
co
r
r
ec
t la
b
els.
Mo
h
it
et
a
l.
[
9
]
co
llected
a
r
ticles
o
n
s
p
o
r
ts
,
h
is
to
r
y
,
s
cien
ce
,
an
d
tech
n
o
lo
g
y
f
r
o
m
W
ik
ip
ed
ia
t
o
cr
ea
te
th
e
Am
er
ican
an
d
Qata
r
i
Mo
d
elin
g
o
f
Ar
ab
ic
(
AQ
MA
R
)
d
ataset
f
o
r
NE
R
m
o
d
el
ass
e
s
s
m
en
t.
T
h
ey
m
an
u
ally
tag
g
ed
th
e
d
ataset
with
th
e
f
o
llo
win
g
tag
s
: p
er
s
o
n
(
PER),
lo
ca
tio
n
(
L
OC
)
,
an
d
o
r
g
an
is
atio
n
(
OR
G)
.
Als
aa
r
an
an
d
Alr
ab
iah
[
6
]
C
r
e
ated
C
ANE
R
C
o
r
p
u
s
,
a
d
ataset
f
o
r
class
ical
A
r
ab
ic
NE
R
b
y
co
llectin
g
258
,
2
6
4
wo
r
d
s
f
r
o
m
th
e
Sah
ih
Al
-
B
u
k
h
a
r
i
b
o
o
k
.
T
h
e
y
u
s
e
2
0
en
tity
ty
p
es,
in
clu
d
in
g
p
er
s
o
n
,
lo
ca
tio
n
,
o
r
g
an
izatio
n
,
Allah
,
p
r
o
p
h
et,
a
n
d
tim
e.
DzN
E
R
is
a
d
atase
t
f
o
r
NE
R
f
o
r
Alg
er
ian
d
ialec
t
co
n
s
tr
u
cte
d
b
y
Dah
o
u
an
d
C
h
er
ag
u
i
[
4
]
.
I
t
co
n
s
is
ts
o
f
2
1
8
3
6
s
en
ten
ce
s
co
llected
f
r
o
m
Yo
u
T
u
b
e
an
d
Face
b
o
o
k
,
an
n
o
tated
to
th
r
ee
en
titi
es:
PE
R
f
o
r
p
er
s
o
n
,
OR
G
f
o
r
o
r
g
an
izatio
n
,
an
d
L
OC
f
o
r
l
o
ca
tio
n
.
An
o
th
er
d
ataset
f
o
r
ANE
R
ca
l
led
W
o
jo
o
d
was
cr
ea
te
d
b
y
J
a
r
r
ar
et
a
l.
[
1
0
]
.
I
t
p
r
o
v
id
es
a
d
ataset
f
o
r
n
ested
Ar
ab
ic
NE
R
co
m
p
r
is
in
g
1
6
,
9
9
9
to
k
en
s
an
d
an
o
th
er
d
ataset
f
o
r
a
f
lat
NE
R
d
atase
t
with
5
8
,
2
7
3
to
k
en
s
.
I
n
b
o
t
h
d
atasets
,
th
e
to
k
en
s
ar
e
tag
g
ed
with
2
1
en
tity
t
y
p
es,
in
clu
d
in
g
PERS
,
DAT
E
,
an
d
L
ANGU
AGE
.
L
iq
r
ein
a
et
a
l
.
[
1
1
]
a
d
d
ed
3
1
s
u
b
ty
p
es
to
W
o
jo
o
d
,
r
esu
ltin
g
in
W
o
jo
o
d
fine,
a
f
in
e
-
g
r
ain
e
d
d
ataset
f
o
r
ANE
R
.
A
f
in
an
cial
NE
R
d
ata
s
et
was
cr
ea
ted
b
y
Ab
d
o
et
a
l.
[
1
2
]
b
y
co
llectin
g
f
i
n
an
cial
a
r
ticles
f
r
o
m
Ar
ab
ic
n
ewsp
ap
er
s
,
r
esu
ltin
g
in
a
d
at
aset c
o
m
p
o
s
ed
o
f
f
in
an
cial
ar
ti
cles f
r
o
m
Ar
ab
ic
n
ewsp
ap
er
s
.
2
.
2
.
Ara
bic
na
m
ed
ent
it
y
mo
dels
Fo
r
th
e
ANE
R
task
,
s
ev
er
al
m
o
d
els
h
av
e
b
ee
n
d
e
v
elo
p
e
d
.
I
n
th
is
s
ec
tio
n
,
we
r
ev
iew
th
e
m
o
s
t
r
ec
en
t
m
eth
o
d
f
o
cu
s
in
g
o
n
th
e
d
ee
p
l
ea
r
n
in
g
a
p
p
r
o
ac
h
es.
T
ab
le
1
s
u
m
m
ar
izes th
ese
ap
p
r
o
ac
h
es.
T
h
e
f
ir
s
t
s
h
ar
ed
task
f
o
r
AN
E
R
was
r
elea
s
ed
b
y
J
ar
r
ar
et
a
l.
[
1
3
]
.
I
n
th
is
task
,
th
e
W
o
jo
o
d
co
r
p
u
s
[
1
3
]
was
u
s
ed
f
o
r
f
lat
an
d
n
ested
ANE
R
[
1
4
]
was
th
e
w
in
n
in
g
team
f
o
r
th
e
NE
R
f
lat
ta
s
k
with
an
F1
-
s
co
r
e
o
f
9
1
.
9
6
%.
W
h
er
ea
s
L
a
o
u
ir
in
e
et
a
l.
[
1
5
]
wo
n
th
e
s
ec
o
n
d
task
b
y
ac
h
iev
in
g
an
F1
Sco
r
e
o
f
9
3
.
7
3
%.
L
iq
r
ein
a
et
a
l
.
[
1
1
]
co
m
p
ar
ed
p
r
etr
ain
ed
lan
g
u
ag
e
m
o
d
els,
in
clu
d
in
g
AR
B
E
R
T
[
1
6
]
,
MA
R
B
E
R
T
v
2
,
an
d
AR
AB
E
R
T
v
2
[
1
7
]
f
o
r
t
h
e
f
in
e
-
g
r
ain
e
d
ANE
R
task
u
s
in
g
th
e
W
o
jo
o
d
f
in
e
d
ataset.
AR
B
E
R
T
ac
h
iev
ed
th
e
h
ig
h
est F1
s
co
r
e
o
f
9
2
%.
AB
io
NE
R
is
a
B
E
R
T
-
b
ased
lan
g
u
ag
e
m
o
d
el
f
o
r
Ar
a
b
ic
b
io
m
ed
ical
NE
R
d
ev
elo
p
ed
b
y
B
o
u
d
jellal
et
a
l.
[
1
8
]
.
T
h
e
y
f
in
e
-
t
u
n
ed
Ar
a
B
E
R
T
[
1
6
]
u
s
in
g
th
e
s
am
e
d
at
aset
as
th
at
u
s
ed
f
o
r
tr
ain
in
g
Ar
aBER
T
,
co
m
b
in
ed
with
m
ed
ical
d
ata
co
llected
f
r
o
m
v
a
r
io
u
s
m
ed
ical
s
o
u
r
ce
s
.
AB
io
NE
R
o
u
tp
er
f
o
r
m
ed
Ar
ab
er
t
an
d
B
E
R
T
m
u
ltil
in
g
u
al
ca
s
es b
y
ac
h
iev
i
n
g
an
F1
Sco
r
e
o
f
8
5
% wh
en
te
s
ted
f
o
r
m
ed
ical
d
ata.
Ar
aBER
T
[
1
6
]
an
d
b
i
d
ir
ec
tio
n
al
g
ated
r
ec
u
r
r
e
n
t
u
n
it
(
B
GR
U)
wer
e
em
p
lo
y
ed
b
y
Als
aa
r
an
an
d
Alr
ab
iah
[
1
9
]
f
o
r
an
ANE
R
m
o
d
el,
wh
ich
was
tr
ain
ed
o
n
th
e
ANE
R
C
o
r
p
d
ataset
m
er
g
ed
with
th
e
AQM
AR
d
ataset,
ac
h
iev
in
g
an
F1
Sco
r
e
o
f
9
0
.
6
8
%.
A
B
iLST
M
-
C
R
F
-
b
ased
m
o
d
el
was
p
r
o
p
o
s
ed
b
y
Me
k
k
i
et
a
l.
[
2
0
]
f
o
r
T
u
n
is
ian
d
ialec
t
NE
R
.
T
h
ey
u
s
ed
t
h
e
T
u
n
is
ian
T
r
ee
B
an
k
co
r
p
u
s
to
tr
ai
n
th
e
m
o
d
el
an
d
T
AD4
6
an
d
T
MD
4
9
co
r
p
o
r
a
to
ev
alu
ate
it.
T
h
is
m
o
d
el
r
ea
c
h
ed
an
F1
-
s
co
r
e
o
f
9
1
.
4
3
%.
Al
-
Qu
r
is
h
i
an
d
So
u
is
s
i
[2
1
]
co
m
b
in
ed
d
if
f
er
e
n
t
lan
g
u
ag
e
m
o
d
els,
in
clu
d
in
g
Ar
aE
L
E
C
T
R
A
[2
2
]
,
Ar
aBER
T
[
1
6
]
,
an
d
XM
L
-
R
o
b
er
ta
with
th
e
r
an
d
o
m
c
o
n
d
itio
n
al
f
ield
(
C
R
F)
alg
o
r
ith
m
.
Fo
r
tr
ain
in
g
,
th
e
y
u
s
ed
AQM
AR
an
d
ANE
R
C
o
r
p
d
atasets
.
Ar
aBER
T
-
C
R
F a
c
h
iev
ed
th
e
b
est r
esu
lts
with
an
ac
cu
r
ac
y
o
f
9
9
%.
AM
W
AL
is
th
e
f
ir
s
t
Ar
ab
ic
NE
R
m
o
d
el
d
esig
n
ed
f
o
r
th
e
f
in
an
cial
d
o
m
ain
,
d
ev
elo
p
e
d
b
y
Ab
d
o
et
a
l.
[
1
2
]
u
s
in
g
Sp
aCy
’
s
b
u
ilt
-
in
NE
R
to
o
lk
it.
T
h
e
m
o
d
el
w
as
tr
ain
ed
o
n
a
d
ataset
tailo
r
ed
f
o
r
th
e
f
i
n
an
cial
f
ield
,
wh
ich
was c
r
ea
ted
b
y
th
e
s
am
e
r
esear
ch
er
s
,
ac
h
iev
in
g
an
F1
s
co
r
e
o
f
9
5
.
9
7
%.
Me
k
k
i
et
a
l.
[
2
0
]
p
r
o
p
o
s
ed
a
B
iLST
M
-
C
R
F
-
b
ased
m
o
d
el
f
o
r
NE
R
in
t
h
e
T
u
n
is
ian
d
i
alec
t.
T
h
e
m
o
d
el
was
tr
ain
e
d
o
n
th
e
m
a
n
u
ally
an
n
o
tated
an
d
POS
-
tag
g
ed
T
u
n
is
i
an
T
r
ee
b
an
k
co
r
p
u
s
.
E
v
alu
atio
n
was
co
n
d
u
cte
d
u
s
in
g
th
e
T
AD4
6
a
n
d
T
MD
4
9
c
o
r
p
o
r
a,
wh
e
r
e
th
e
m
o
d
el
ac
h
ie
v
ed
an
F1
s
co
r
e
o
f
9
1
.
4
3
%.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
42
,
No
.
1
,
Ap
r
il
20
26
:
1
42
-
14
8
144
Als
aa
r
an
an
d
Alr
ab
iah
[
1
9
]
d
ev
elo
p
ed
a
d
ee
p
lear
n
in
g
m
o
d
el
f
o
r
NE
R
in
class
ica
l
Ar
ab
ic
u
s
in
g
th
e
C
ANE
R
C
o
r
p
u
s
.
T
h
e
m
o
d
el
l
ev
er
ag
es
th
e
Ar
aBER
T
p
r
e
-
tr
ain
ed
lan
g
u
a
g
e
m
o
d
el
to
ex
tr
ac
t
co
n
tex
tu
alize
d
r
ep
r
esen
tatio
n
s
o
f
th
e
in
p
u
t
t
ex
t.
I
t
co
m
b
in
es
B
iLST
M,
B
GR
U,
an
d
C
R
F
lay
er
s
,
ac
h
iev
in
g
an
F1
s
co
r
e
o
f
9
4
.
7
6
%.
I
n
an
o
th
er
s
tu
d
y
,
Salah
et
a
l.
[
2
3
]
ad
o
p
ted
an
ML
ap
p
r
o
ac
h
f
o
r
class
ical
Ar
ab
i
c
NE
R
,
wh
er
e
th
ey
em
p
lo
y
ed
NB
,
wh
ich
ac
h
iev
e
d
an
F1
s
co
r
e
o
f
80%
.
T
ab
le
1
.
Su
m
m
a
r
y
o
f
Ar
ab
ic
NE
R
m
o
d
els
R
e
f
e
r
e
n
c
e
D
a
t
a
s
e
t
M
o
d
e
l
Ev
a
l
u
a
t
i
o
n
[
1
1
]
W
o
j
o
o
d
f
i
n
e
A
R
B
E
R
T
F1
-
sc
o
r
e
=
9
2
%
[
1
8
]
A
med
i
c
a
l
d
a
t
a
set
A
r
a
B
ER
T
F1
-
sc
o
r
e
=
8
5
%
[
1
9
]
A
N
ER
C
o
r
p
,
A
Q
M
A
R
A
r
a
B
ER
T+
B
G
R
U
F1
-
sc
o
r
e
=
9
0
.
6
8
%.
[
2
0
]
Tr
e
e
B
a
n
k
,
TA
D
4
6
,
TM
D
4
9
B
i
LST
M
-
CRF
F1
-
sc
o
r
e
=
9
1
.
4
3
%
[2
1
]
A
Q
M
A
R
,
A
N
E
R
C
o
r
p
A
r
a
B
ER
T
-
C
R
F
A
c
c
u
r
a
c
y
=
9
9
%
[
1
2
]
F
i
n
a
n
c
i
a
l
d
a
t
a
s
e
t
s
S
p
a
C
y
’
s NE
R
t
o
o
l
k
i
t
F1
-
sc
o
r
e
=
9
5
.
9
7
%.
3.
M
AT
E
R
I
AL
S AN
D
M
E
T
H
O
DS
I
n
th
is
s
ec
tio
n
,
we
d
escr
ib
e
th
e
ar
ch
itectu
r
e
o
f
th
e
p
r
o
p
o
s
ed
m
o
d
el
an
d
th
e
d
ataset
u
s
ed
.
3
.
1
.
Da
t
a
s
et
I
n
th
is
r
esear
ch
,
we
ch
o
s
e
to
wo
r
k
o
n
th
e
W
o
jo
o
d
NE
R
d
a
taset
f
o
r
s
ev
er
al
r
ea
s
o
n
s
.
Firs
tly
,
it
is
a
r
ec
en
t
lar
g
e
d
ataset
co
m
p
o
s
ed
o
f
2
6
k
s
en
ten
ce
s
an
d
5
5
0
k
to
k
en
s
.
Seco
n
d
ly
,
th
e
W
o
j
o
o
d
d
ataset
is
s
tr
u
ctu
r
ed
f
o
r
two
ty
p
es
o
f
NE
R
:
f
lat
NE
R
an
d
n
ested
NE
R
.
F
in
ally
,
th
e
to
k
en
s
ar
e
an
n
o
tated
o
n
2
1
en
tity
ty
p
es,
p
r
o
v
id
i
n
g
a
r
ic
h
an
d
f
in
e
-
g
r
ain
ed
s
em
an
tic
r
ep
r
esen
tatio
n
o
f
s
em
an
tic
i
n
f
o
r
m
atio
n
,
wh
ich
is
cr
u
cial
f
o
r
tr
ain
in
g
a
co
m
p
r
e
h
en
s
iv
e
m
o
d
el,
esp
ec
ially
f
o
r
lan
g
u
ag
es wi
th
r
ich
m
o
r
p
h
o
lo
g
y
,
s
u
ch
as A
r
ab
ic.
T
o
cr
ea
te
Wo
jo
o
d
,
J
ar
r
ar
et
a
l.
[
1
0
]
co
llected
ar
ticles
f
r
o
m
d
iv
er
s
e
s
o
u
r
ce
s
a
n
d
v
ar
io
u
s
t
o
p
ics.
T
h
e
s
ec
o
n
d
p
a
r
t
o
f
th
e
d
ataset
was
ex
tr
ac
ted
f
r
o
m
B
ir
ze
it
Un
iv
er
s
ity
'
s
ar
ch
iv
e.
I
n
ad
d
i
tio
n
to
a
d
ataset
o
f
Palest
in
ian
d
ialec
ts
[
2
4
]
an
d
a
d
ataset
o
f
L
eb
an
ese
d
ialec
t
[
2
5
]
.
W
o
jo
o
d
is
co
m
p
o
s
ed
o
f
5
5
0
k
to
k
e
n
s
m
an
u
ally
an
n
o
tated
in
t
o
2
1
en
tity
ty
p
es d
escr
ib
ed
in
T
ab
le
2
,
an
d
T
a
b
le
3
r
e
p
r
esen
ts
th
e
s
ize
o
f
th
e
d
ataset.
T
ab
le
2
.
W
o
jo
o
d
en
tity
ty
p
es d
escr
ip
tio
n
T
ab
le
3
.
W
o
jo
o
d
NE
R
d
ataset
s
ize
Tr
a
i
n
i
n
g
se
t
D
e
v
s
e
t
Te
st
set
D
a
t
a
s
e
t
s
i
z
e
(
se
n
t
e
n
c
e
s)
2
3
1
2
5
3
3
0
4
6
6
0
6
3
.
2
.
M
o
del a
rc
hite
ct
ure
Fo
r
th
e
Ar
ab
ic
n
am
ed
en
tity
r
ec
o
g
n
itio
n
,
we
tr
ain
d
if
f
e
r
en
t
m
o
d
els
o
n
th
e
W
o
jo
o
d
NE
R
f
lat
tr
ain
d
ataset.
I
t is
co
m
p
o
s
ed
o
f
5
5
0
k
to
k
en
s
an
n
o
tated
in
t
o
2
1
en
t
ity
ty
p
es.
T
h
e
f
ir
s
t
m
o
d
el
is
a
co
m
b
in
a
tio
n
o
f
th
e
Ar
aE
L
E
C
T
R
A
lan
g
u
ag
e
m
o
d
el
with
t
h
e
B
iLST
M
n
eu
r
al
n
etwo
r
k
.
W
e
u
s
e
Ar
aE
L
E
C
T
R
A
to
ex
tr
ac
t
a
c
o
n
tex
tu
alize
d
r
ep
r
esen
tatio
n
f
o
r
ea
ch
to
k
en
as
a
v
ec
to
r
o
f
s
h
a
p
e
(
5
0
,
7
6
8
)
,
w
h
er
e
5
0
is
th
e
m
a
x
im
u
m
s
eq
u
en
ce
le
n
g
th
an
d
7
6
8
is
th
e
h
id
d
en
s
ize
in
th
e
Ar
aE
L
E
C
T
R
A
m
o
d
el.
T
h
is
v
ec
to
r
is
th
en
f
ed
in
to
a
B
iLST
M
lay
er
o
f
1
2
8
u
n
its
to
ca
p
tu
r
e
b
id
ir
ec
tio
n
al
co
n
tex
tu
al
d
ep
en
d
en
cies in
a
s
eq
u
en
ce
.
Fin
ally
,
a
s
o
f
tm
ax
d
en
s
e
lay
er
is
ap
p
lied
f
o
r
class
if
icatio
n
.
T
h
e
ar
ch
itectu
r
e
o
f
th
is
m
o
d
el
is
d
ep
icted
in
Fig
u
r
e
1
.
I
n
th
e
s
ec
o
n
d
m
o
d
el,
we
r
ep
lace
th
e
B
iLST
M
with
B
GR
U
n
etwo
r
k
s
.
T
h
e
f
in
al
m
o
d
el
is
a
co
m
b
in
atio
n
o
f
Ar
aE
L
E
C
T
R
A,
a
co
n
v
o
l
u
tio
n
al
n
eu
r
al
n
et
wo
r
k
(
C
NN)
,
an
d
B
iLST
M.
Ar
aE
L
E
C
T
R
A
is
u
s
ed
Ta
g
D
e
scri
p
t
i
o
n
Ta
g
D
e
scri
p
t
i
o
n
P
ER
S
P
e
o
p
l
e
n
a
me
s
D
A
TE
S
p
e
c
i
f
i
c
o
r
r
e
l
a
t
i
v
e
d
a
t
e
s
N
O
R
P
G
r
o
u
p
o
f
p
e
o
p
l
e
TI
M
E
S
p
e
c
i
f
i
c
o
r
r
e
l
a
t
i
v
e
t
i
m
e
s
O
C
C
O
c
c
u
p
a
t
i
o
n
LA
N
G
U
A
G
E
N
a
med
h
u
m
a
n
l
a
n
g
u
a
g
e
O
R
G
O
r
g
a
n
i
z
a
t
i
o
n
s
W
EB
S
I
TE
N
a
me
o
f
w
e
b
si
t
e
G
P
E
G
e
o
p
o
l
i
t
i
c
a
l
l
i
k
e
c
o
u
n
t
r
i
e
s
LA
W
R
e
f
e
r
e
n
c
e
t
o
a
s
p
e
c
i
f
i
c
l
a
w
LO
C
G
e
o
g
r
a
p
h
i
c
a
l
l
o
c
a
t
i
o
n
s
C
A
R
D
I
N
A
L
N
u
mera
l
s wr
i
t
t
e
n
i
n
d
i
g
i
t
s
o
r
w
o
r
d
s
F
A
C
N
a
mes
o
f
s
p
e
c
i
f
i
c
p
l
a
c
e
s
O
R
D
I
N
A
L
O
r
d
i
n
a
l
n
u
m
b
e
r
P
R
O
D
U
C
T
N
a
mes
o
f
p
r
o
d
u
c
t
s
P
ER
C
EN
T
A
w
o
r
d
o
r
a
s
y
m
b
o
l
r
e
f
e
r
r
i
n
g
t
o
a
p
e
r
c
e
n
t
EV
EN
T
N
a
me
o
f
a
n
e
v
e
n
t
Q
U
A
N
T
I
TY
v
a
l
u
e
m
e
a
s
u
r
e
d
b
y
st
a
n
d
a
r
d
i
z
e
d
u
n
i
t
s
C
U
R
R
A
n
y
n
a
me
o
r
s
y
m
b
o
l
r
e
f
e
r
r
i
n
g
t
o
c
u
r
r
e
n
c
y
.
U
N
I
T
N
a
mes
o
f
s
t
a
n
d
a
r
d
i
z
e
d
u
n
i
t
s
M
O
N
EY
A
b
so
l
u
t
e
mo
n
e
t
a
r
y
q
u
a
n
t
i
t
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
La
n
g
u
a
g
e
mo
d
els a
n
d
d
ee
p
n
e
u
r
a
l n
etw
o
r
ks fo
r
A
r
a
b
ic
n
a
med
en
tity reco
g
n
itio
n
(
S
o
mia
K
h
ed
imi
)
145
f
o
r
wo
r
d
em
b
ed
d
in
g
,
wh
ile
C
NN
h
elp
s
ca
p
tu
r
e
l
o
ca
l
p
att
er
n
s
an
d
s
h
o
r
t
-
ter
m
d
e
p
en
d
e
n
cies,
an
d
B
iLST
M
m
o
d
els lo
n
g
-
r
an
g
e
an
d
b
id
ir
ec
tio
n
al
co
n
tex
t.
Fig
u
r
e
1
.
Ar
aE
L
E
C
T
R
A
-
B
iL
STM
m
o
d
el
ar
ch
itectu
r
e
4.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
I
n
th
is
s
ec
tio
n
,
we
p
r
esen
t
th
e
o
b
tain
ed
r
esu
lts
o
f
ea
ch
m
o
d
el.
W
e
s
e
lect
1
0
0
0
ex
am
p
le
s
f
r
o
m
th
e
W
o
jo
o
d
NE
R
test
d
atasets
to
ev
alu
ate
ea
ch
m
o
d
el.
T
ab
le
4
r
ep
r
esen
ts
th
e
test
ac
cu
r
ac
y
ac
h
iev
ed
b
y
ea
c
h
m
o
d
el.
T
h
e
B
iLST
M
m
o
d
el
o
u
tp
er
f
o
r
m
s
th
e
o
th
er
m
o
d
els,
ac
h
iev
in
g
an
ac
c
u
r
ac
y
o
f
9
8
.
2
9
%.
wh
ile
th
e
B
GR
U
an
d
C
NN
-
B
iLST
M
a
ch
iev
e
9
7
.
7
0
%
an
d
9
7
.
9
7
%
o
f
ac
cu
r
ac
y
,
r
esp
ec
tiv
el
y
.
On
th
e
o
th
er
h
an
d
,
B
iLST
M
ac
h
iev
es
8
6
%,
8
8
%,
an
d
8
7
%
p
r
ec
is
io
n
,
r
ec
al
l,
an
d
F1
-
s
co
r
e,
r
esp
ec
tiv
ely
,
o
u
tp
er
f
o
r
m
in
g
th
e
B
GR
U
an
d
C
NN
-
B
iLST
M
m
o
d
e
ls
,
as d
ep
icted
in
Fig
u
r
e
2
.
T
ab
le
4
.
T
est ac
cu
r
ac
y
f
o
r
ea
c
h
m
o
d
el
M
o
d
e
l
BiL
S
TM
BG
RU
CNN
-
BiL
S
TM
Tes
t
a
c
c
u
ra
c
y
98
.
2
9
%
97
.
7
0
%
97
.
9
7
%
I
n
n
am
e
d
en
tity
r
ec
o
g
n
itio
n
,
th
e
co
n
tex
t
b
ef
o
r
e
an
d
af
ter
a
wo
r
d
is
cr
u
cial
to
id
en
tif
y
i
ts
ty
p
e;
in
ad
d
itio
n
,
th
e
m
ea
n
in
g
o
f
a
wo
r
d
d
ep
en
d
s
o
n
b
o
t
h
d
is
tan
t
an
d
n
ea
r
b
y
c
o
n
tex
t.
Fo
r
e
x
am
p
l
e,
in
th
e
s
en
ten
ce
“
ق
ب
اس
ير
صم
سي
ئ
ر
كر
اب
م
ينسح
“,
th
e
wo
r
d
“
كر
اب
م
“
is
a
f
i
r
s
t
n
am
e
o
f
a
p
er
s
o
n
,
n
o
t
th
e
ad
jectiv
e
‘
b
less
ed
’
.
Fo
r
th
is
r
ea
s
o
n
,
we
n
ee
d
a
m
o
d
el
th
at
ca
n
ca
p
tu
r
e
th
e
lo
n
g
an
d
s
h
o
r
t
-
ter
m
d
ep
en
d
en
cies
b
etwe
en
th
e
wo
r
d
s
.
On
th
e
o
th
er
h
a
n
d
,
b
o
th
L
STM
an
d
GR
U
h
av
e
m
em
o
r
y
ce
lls
an
d
g
ates,
b
u
t
th
e
L
STM
'
s
m
em
o
r
y
is
m
o
r
e
c
o
m
p
lex
th
an
th
e
GR
U’
s
m
em
o
r
y
,
wh
i
ch
allo
ws th
em
to
ca
p
tu
r
e
lo
n
g
-
ter
m
d
e
p
en
d
e
n
cies b
etter
th
an
th
e
GR
U.
Alth
o
u
g
h
C
NNs
ar
e
u
s
ef
u
l
f
o
r
ex
tr
ac
tin
g
lo
ca
l
p
atter
n
s
,
a
s
tan
d
alo
n
e
B
iLST
M
p
er
f
o
r
m
s
b
etter
th
an
th
e
co
m
b
in
atio
n
o
f
C
NN
-
B
iLST
M
f
o
r
NE
R
.
C
NN
ap
p
lies
f
ilter
s
,
g
en
er
ally
s
m
all
2
-
5
,
o
v
e
r
win
d
o
ws
o
f
to
k
en
s
,
wh
ich
ca
n
m
o
d
if
y
th
e
s
tr
u
ctu
r
e
o
f
th
e
s
eq
u
en
ce
,
ca
u
s
in
g
lo
s
s
o
f
in
f
o
r
m
atio
n
.
I
n
co
n
tr
ast,
B
iLST
M
p
r
o
ce
s
s
es
th
e
s
eq
u
en
ce
to
k
en
b
y
to
k
en
in
b
o
th
d
ir
ec
tio
n
s
,
ca
p
tu
r
in
g
L
o
n
g
-
r
an
g
e
d
e
p
en
d
en
cies
wh
ich
C
NN
o
f
ten
m
is
s
es.
I
n
T
ab
le
5
,
we
co
m
p
a
r
e
o
u
r
m
o
d
el
with
th
e
p
r
ev
io
u
s
p
r
o
p
o
s
ed
m
o
d
els
f
o
r
ANE
R
u
s
in
g
th
e
W
o
jo
o
d
NE
R
d
ataset.
[
1
4
]
a
d
d
r
ess
ed
th
e
NE
R
p
r
o
b
le
m
as
a
s
p
an
class
if
icatio
n
p
r
o
b
lem
,
wh
er
e
th
e
y
u
s
ed
Ar
aBER
T
f
o
llo
wed
b
y
B
iLST
M
f
o
r
to
k
e
n
r
ep
r
esen
tatio
n
.
W
h
er
ea
s
[
1
5
]
r
elied
o
n
Dif
f
u
s
io
n
NE
R
an
d
Par
allel
I
n
s
tan
ce
Qu
er
y
Netwo
r
k
m
o
d
el
ap
p
r
o
ac
h
u
s
in
g
B
E
R
T
-
b
ased
m
o
d
els.
T
h
ese
m
o
d
els
o
u
tp
e
r
f
o
r
m
e
d
o
u
r
m
o
d
el
b
ec
au
s
e
th
ey
m
o
d
el
en
titi
es
as
s
tr
u
ctu
r
ed
o
b
jects
with
ex
p
licit
b
o
u
n
d
ar
ies
an
d
g
l
o
b
al
in
t
er
ac
tio
n
s
,
wh
er
ea
s
o
u
r
m
o
d
el
class
if
ies to
k
en
s
in
d
ep
en
d
e
n
tly
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
42
,
No
.
1
,
Ap
r
il
20
26
:
1
42
-
14
8
146
Fig
u
r
e
2
.
E
v
alu
atio
n
o
f
NE
R
m
o
d
els
T
ab
le
5
.
NE
R
m
o
d
els co
m
p
ar
i
s
o
n
.
M
o
d
e
l
P
r
e
c
i
s
i
o
n
R
e
c
a
l
l
F1
-
sc
o
r
e
[
1
4
]
9
2
.
5
6
%
9
1
.
3
6
%
9
1
.
9
6
%
[
1
5
]
9
1
.
9
2
%
9
1
.
8
8
%
9
1
.
9
6
%
O
u
r
8
6
%
8
8
%
8
7
%
5.
CO
NCLU
SI
O
N
T
h
i
s
s
tu
d
y
p
r
esen
ted
a
d
ee
p
lear
n
in
g
ap
p
r
o
ac
h
f
o
r
Ar
ab
ic
NE
R
lev
er
ag
in
g
t
h
e
Ar
aE
L
E
C
T
R
A
lan
g
u
ag
e
m
o
d
el
an
d
th
e
B
iLST
M
n
eu
r
al
n
etwo
r
k
.
W
e
u
s
ed
th
e
W
o
jo
o
d
NE
R
d
ataset,
wh
ich
is
a
f
in
e
-
g
r
ain
e
d
d
ataset
th
at
p
r
o
v
id
es
d
etailed
an
n
o
tatio
n
s
ac
r
o
s
s
2
1
en
tity
ty
p
es,
wh
ich
is
cr
u
cial
f
o
r
d
ev
el
o
p
in
g
a
m
o
d
el
th
at
ca
n
ac
cu
r
ately
d
is
tin
g
u
is
h
b
e
twee
n
s
u
b
tle
ca
teg
o
r
ies,
s
u
ch
as
d
if
f
e
r
en
t
t
y
p
es
o
f
Nu
m
b
e
r
s
(
DAT
E
,
T
I
ME
,
C
AR
DI
NAL
,
OR
DI
NAL
)
.
T
h
e
co
m
b
i
n
atio
n
o
f
Ar
aE
L
E
C
T
R
A
an
d
B
iLST
M
f
o
r
th
e
NE
R
task
h
as
p
r
o
v
en
its
ef
f
i
cien
cy
b
y
ac
h
iev
i
ng
an
ac
cu
r
ac
y
o
f
9
8
.
2
9
%
an
d
a
n
F1
-
s
co
r
e
o
f
8
7
%
o
u
tp
er
f
o
r
m
in
g
th
e
B
GR
U
an
d
C
NN
-
B
iLST
M
m
o
d
els.
T
h
ese
p
r
o
m
is
in
g
r
e
s
u
lts
ca
n
h
elp
in
d
ev
elo
p
i
n
g
r
o
b
u
s
t
n
atu
r
al
lan
g
u
ag
e
ap
p
licatio
n
s
,
s
u
ch
as
q
u
esti
o
n
-
an
s
wer
in
g
s
y
s
tem
s
,
t
ex
t su
m
m
ar
izatio
n
,
a
n
d
m
ac
h
i
n
e
tr
an
s
latio
n
.
H
o
w
e
v
e
r
,
s
p
an
-
b
ased
NE
R
b
ec
o
m
es
v
er
y
im
p
o
r
tan
t
d
u
e
to
th
e
r
ich
n
ess
o
f
Ar
ab
ic
tex
t,
wh
er
e
m
u
ltip
le
en
titi
es
m
ay
b
e
n
est
ed
with
in
th
e
s
am
e
tex
tu
al
s
p
an
.
Fu
t
u
r
e
wo
r
k
will
ex
p
lo
r
e
m
o
r
e
ex
p
r
ess
iv
e
m
o
d
elin
g
p
ar
ad
i
g
m
s
f
o
r
Ar
a
b
ic
n
ested
n
am
e
d
en
tity
r
ec
o
g
n
itio
n
,
s
u
ch
as
s
p
an
-
b
ased
,
q
u
er
y
-
b
ased
,
an
d
d
if
f
u
s
io
n
-
b
ased
ap
p
r
o
ac
h
es,
t
o
b
etter
ca
p
tu
r
e
o
v
er
lap
p
in
g
an
d
h
ier
ar
c
h
ically
n
ested
e
n
titi
es.
I
n
ad
d
itio
n
,
lev
er
ag
in
g
lar
g
er
Ar
a
b
ic
p
r
etr
ain
ed
lan
g
u
a
g
e
m
o
d
els
an
d
e
v
alu
atin
g
th
ese
m
eth
o
d
s
ac
r
o
s
s
d
if
f
er
en
t
d
o
m
ain
s
an
d
lan
g
u
ag
e
v
ar
ieties r
em
ain
im
p
o
r
tan
t d
i
r
ec
tio
n
s
f
o
r
f
u
r
t
h
e
r
r
esear
ch
.
F
UNDING
I
NF
O
R
M
A
T
I
O
N
T
h
is
r
esear
ch
r
ec
eiv
ed
n
o
e
x
te
r
n
al
f
u
n
d
in
g
.
AUTHO
R
CO
NT
RI
B
UT
I
O
NS ST
A
T
E
M
E
N
T
T
h
is
jo
u
r
n
al
u
s
es
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT)
to
r
ec
o
g
n
ize
in
d
iv
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
th
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
llab
o
r
atio
n
.
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
So
m
ia
Kh
ed
im
i
✓
✓
✓
✓
✓
✓
✓
✓
✓
Ab
d
elg
h
an
i Bo
u
zia
n
e
✓
✓
✓
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
I
SS
N:
2
5
0
2
-
4
7
52
La
n
g
u
a
g
e
mo
d
els a
n
d
d
ee
p
n
e
u
r
a
l n
etw
o
r
ks fo
r
A
r
a
b
ic
n
a
med
en
tity reco
g
n
itio
n
(
S
o
mia
K
h
ed
imi
)
147
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
s
i
s
I
:
I
n
v
e
s
t
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
s
i
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
T
h
e
a
u
th
o
r
s
s
tate
n
o
co
n
f
lict o
f
in
ter
est.
DATA AV
AI
L
AB
I
L
I
T
Y
-
W
OJO
OD
d
ataset
was
d
o
wn
lo
ad
ed
f
r
o
m
h
ttp
s
://s
in
a.
b
ir
ze
it.e
d
u
/wo
jo
o
d
/
o
r
ch
ec
k
h
ttp
s
://g
ith
u
b
.
co
m
/Si
n
aL
ab
/Ar
ab
icNE
R
.
-
T
h
e
Py
th
o
n
c
o
d
e
an
d
ev
alu
atio
n
f
o
r
ea
c
h
m
o
d
el
ar
e
av
ailab
le
o
n
h
ttp
s
://g
ith
u
b
.
co
m
/ai
-
p
h
d
-
d
r
ea
m
/ANE
R
/tre
e/m
ain
RE
F
E
R
E
NC
E
S
[
1
]
S
.
X
i
e
,
Y
.
X
i
a
,
L.
W
u
,
Y
.
H
u
a
n
g
,
Y
.
F
a
n
,
a
n
d
T
.
Q
i
n
,
“
E
n
d
-
to
-
e
n
d
e
n
t
i
t
y
-
a
w
a
r
e
n
e
u
r
a
l
ma
c
h
i
n
e
t
r
a
n
s
l
a
t
i
o
n
,
”
Ma
c
h
i
n
e
L
e
a
rn
i
n
g
,
v
o
l
.
1
1
1
,
n
o
.
3
,
p
p
.
1
1
8
1
–
1
2
0
3
,
M
a
r
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
0
9
9
4
-
0
2
1
-
0
6
0
7
3
-
9.
[
2
]
B
.
L
i
u
,
H
.
W
e
i
,
D
.
N
i
u
,
H
.
C
h
e
n
,
a
n
d
Y
.
H
e
,
“
A
sk
i
n
g
q
u
e
s
t
i
o
n
s
t
h
e
h
u
m
a
n
w
a
y
:
s
c
a
l
a
b
l
e
q
u
e
st
i
o
n
-
a
n
sw
e
r
g
e
n
e
r
a
t
i
o
n
f
r
o
m
t
e
x
t
c
o
r
p
u
s
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
T
h
e
W
e
b
C
o
n
f
e
re
n
c
e
2
0
2
0
,
N
e
w
Y
o
r
k
,
N
Y
,
U
S
A
:
A
C
M
,
A
p
r
.
2
0
2
0
,
p
p
.
2
0
3
2
–
2
0
4
3
.
d
o
i
:
1
0
.
1
1
4
5
/
3
3
6
6
4
2
3
.
3
3
8
0
2
7
0
.
[
3
]
S
.
B
e
r
e
z
i
n
a
n
d
T
.
B
a
t
u
r
a
,
“
N
a
me
d
e
n
t
i
t
y
i
n
c
l
u
s
i
o
n
i
n
a
b
s
t
r
a
c
t
i
v
e
t
e
x
t
s
u
m
mariz
a
t
i
o
n
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
T
h
i
rd
W
o
rks
h
o
p
o
n
S
c
h
o
l
a
rl
y
D
o
c
u
m
e
n
t
Pr
o
c
e
ss
i
n
g
,
2
0
2
2
,
p
p
.
1
5
8
–
1
6
2
.
[
4
]
A
.
H
.
D
a
h
o
u
a
n
d
M
.
A
.
C
h
e
r
a
g
u
i
,
“
D
z
N
ER
:
a
l
a
r
g
e
A
l
g
e
r
i
a
n
n
a
me
d
e
n
t
i
t
y
r
e
c
o
g
n
i
t
i
o
n
d
a
t
a
se
t
,
”
N
a
t
u
r
a
l
L
a
n
g
u
a
g
e
Pr
o
c
e
ss
i
n
g
J
o
u
rn
a
l
,
v
o
l
.
3
,
p
.
1
0
0
0
0
5
,
J
u
n
.
2
0
2
3
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
n
l
p
.
2
0
2
3
.
1
0
0
0
0
5
.
[
5
]
O
.
H
.
Za
y
e
d
a
n
d
S
.
R
.
E
l
-
B
e
l
t
a
g
y
,
“
N
a
med
e
n
t
i
t
y
r
e
c
o
g
n
i
t
i
o
n
o
f
p
e
r
so
n
s
’
n
a
me
s
i
n
A
r
a
b
i
c
t
w
e
e
t
s,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
R
e
c
e
n
t
Ad
v
a
n
c
e
s i
n
N
a
t
u
ra
l
L
a
n
g
u
a
g
e
Pr
o
c
e
ssi
n
g
,
2
0
1
5
,
p
p
.
7
3
1
–
7
3
8
.
[
6
]
N
.
A
l
saa
r
a
n
a
n
d
M
.
A
l
r
a
b
i
a
h
,
“
C
l
a
ss
i
c
a
l
A
r
a
b
i
c
n
a
me
d
e
n
t
i
t
y
r
e
c
o
g
n
i
t
i
o
n
u
si
n
g
v
a
r
i
a
n
t
d
e
e
p
n
e
u
r
a
l
n
e
t
w
o
r
k
a
r
c
h
i
t
e
c
t
u
r
e
s
a
n
d
B
ER
T,
”
I
EEE
Ac
c
e
ss
,
v
o
l
.
9
,
p
p
.
9
1
5
3
7
–
9
1
5
4
7
,
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
A
C
C
E
S
S
.
2
0
2
1
.
3
0
9
2
2
6
1
.
[
7
]
Y
.
B
e
n
a
j
i
b
a
,
M
.
D
i
a
b
,
a
n
d
P
.
R
o
ss
o
,
“
A
r
a
b
i
c
n
a
me
d
e
n
t
i
t
y
r
e
c
o
g
n
i
t
i
o
n
u
si
n
g
o
p
t
i
mi
z
e
d
f
e
a
t
u
r
e
se
t
s,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
C
o
n
f
e
re
n
c
e
o
n
Em
p
i
r
i
c
a
l
Me
t
h
o
d
s
i
n
N
a
t
u
r
a
l
L
a
n
g
u
a
g
e
Pr
o
c
e
ssi
n
g
-
EM
N
L
P
’
0
8
,
M
o
r
r
i
st
o
w
n
,
N
J,
U
S
A
:
A
sso
c
i
a
t
i
o
n
f
o
r
C
o
m
p
u
t
a
t
i
o
n
a
l
L
i
n
g
u
i
s
t
i
c
s,
2
0
0
8
,
p
.
2
8
4
.
d
o
i
:
1
0
.
3
1
1
5
/
1
6
1
3
7
1
5
.
1
6
1
3
7
5
5
.
[
8
]
M
.
A
l
-
D
u
w
a
i
s
,
H
.
A
l
-
K
h
a
l
i
f
a
,
a
n
d
A
.
A
l
-
S
a
l
ma
n
,
“
C
L
EA
N
A
N
ER
C
o
r
p
:
i
d
e
n
t
i
f
y
i
n
g
a
n
d
c
o
r
r
e
c
t
i
n
g
i
n
c
o
r
r
e
c
t
l
a
b
e
l
s
i
n
t
h
e
A
N
ER
c
o
r
p
d
a
t
a
se
t
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
6
t
h
Wo
rks
h
o
p
o
n
O
p
e
n
-
S
o
u
rc
e
Ar
a
b
i
c
C
o
r
p
o
r
a
a
n
d
Pro
c
e
ss
i
n
g
T
o
o
l
s
(
O
S
AC
T
)
w
i
t
h
S
h
a
re
d
T
a
sk
s
o
n
Ar
a
b
i
c
L
L
Ms
H
a
l
l
u
c
i
n
a
t
i
o
n
a
n
d
D
i
a
l
e
c
t
t
o
MS
A
Ma
c
h
i
n
e
T
ra
n
s
l
a
t
i
o
n
@
L
R
EC
-
C
O
L
I
N
G
2
0
2
4
,
2
0
2
4
,
p
p
.
1
3
–
1
9
.
[
9
]
B
.
M
o
h
i
t
,
N
.
S
c
h
n
e
i
d
e
r
,
R
.
B
h
o
w
m
i
c
k
,
K
.
O
f
l
a
z
e
r
,
a
n
d
N
.
A
.
S
m
i
t
h
,
“
R
e
c
a
l
l
-
o
r
i
e
n
t
e
d
l
e
a
r
n
i
n
g
o
f
n
a
me
d
e
n
t
i
t
i
e
s
i
n
A
r
a
b
i
c
W
i
k
i
p
e
d
i
a
,
”
i
n
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
1
3
t
h
C
o
n
f
e
r
e
n
c
e
o
f
t
h
e
E
u
ro
p
e
a
n
C
h
a
p
t
e
r
o
f
t
h
e
Ass
o
c
i
a
t
i
o
n
f
o
r
C
o
m
p
u
t
a
t
i
o
n
a
l
L
i
n
g
u
i
st
i
c
s
,
2
0
1
2
,
p
p
.
1
6
2
–
1
7
3
.
[
1
0
]
M
.
J
a
r
r
a
r
,
M
.
K
h
a
l
i
l
i
a
,
a
n
d
S
.
G
h
a
n
e
m,
“
W
o
j
o
o
d
:
n
e
st
e
d
a
r
a
b
i
c
n
a
m
e
d
e
n
t
i
t
y
c
o
r
p
u
s
a
n
d
r
e
c
o
g
n
i
t
i
o
n
u
si
n
g
B
ER
T,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
T
h
i
rt
e
e
n
t
h
L
a
n
g
u
a
g
e
Re
s
o
u
rce
s
a
n
d
E
v
a
l
u
a
t
i
o
n
C
o
n
f
e
re
n
c
e
,
2
0
2
2
,
p
p
.
3
6
2
6
–
3
6
3
6
.
[
1
1
]
H
.
Li
q
r
e
i
n
a
,
M
.
J
a
r
r
a
r
,
M
.
K
h
a
l
i
l
i
a
,
A
.
El
-
S
h
a
n
g
i
t
i
,
a
n
d
M
.
A
b
d
u
l
-
M
a
g
e
e
d
,
“
A
r
a
b
i
c
f
i
n
e
-
g
r
a
i
n
e
d
e
n
t
i
t
y
r
e
c
o
g
n
i
t
i
o
n
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
Ara
b
i
c
N
L
P
2
0
2
3
,
S
t
r
o
u
d
s
b
u
r
g
,
P
A
,
U
S
A
:
A
ss
o
c
i
a
t
i
o
n
f
o
r
C
o
mp
u
t
a
t
i
o
n
a
l
Li
n
g
u
i
st
i
c
s,
2
0
2
3
,
p
p
.
3
1
0
–
3
2
3
.
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
2
0
2
3
.
a
r
a
b
i
c
n
l
p
-
1
.
2
5
.
[
1
2
]
M
.
S
.
A
b
d
o
,
Y
.
H
a
t
e
k
a
r
,
a
n
d
D
.
C
a
v
a
r
,
“
A
M
W
A
L
:
n
a
me
d
e
n
t
i
t
y
r
e
c
o
g
n
i
t
i
o
n
f
o
r
A
r
a
b
i
c
f
i
n
a
n
c
i
a
l
n
e
w
s,
”
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
J
o
i
n
t
Wo
r
k
sh
o
p
o
f
t
h
e
9
t
h
F
i
n
a
n
c
i
a
l
T
e
c
h
n
o
l
o
g
y
a
n
d
N
a
t
u
ra
l
L
a
n
g
u
a
g
e
Pro
c
e
ss
i
n
g
(
F
i
n
N
L
P)
,
t
h
e
6
t
h
F
i
n
a
n
c
i
a
l
N
a
rr
a
t
i
v
e
Pr
o
c
e
s
si
n
g
(
FN
P)
,
a
n
d
t
h
e
1
s
t
Wo
r
k
sh
o
p
o
n
L
a
r
g
e
L
a
n
g
u
a
g
e
M
o
d
e
l
s
f
o
r F
i
n
a
n
c
e
a
n
d
L
e
g
a
l
(
L
L
MF
i
n
L
e
g
a
l
)
,
p
p
.
2
0
7
–
2
1
3
,
2
0
2
5
.
[
1
3
]
M
.
J
a
r
r
a
r
e
t
a
l
.
,
“
W
o
j
o
o
d
N
ER
2
0
2
3
:
t
h
e
f
i
r
st
a
r
a
b
i
c
n
a
me
d
e
n
t
i
t
y
r
e
c
o
g
n
i
t
i
o
n
s
h
a
r
e
d
t
a
s
k
,
”
Pr
o
c
e
e
d
i
n
g
s
o
f
Ar
a
b
i
c
N
L
P
,
p
p
.
7
4
8
–
7
5
8
,
2
0
2
3
.
[
1
4
]
N
.
E
l
k
h
b
i
r
,
U
.
Za
r
a
t
i
a
n
a
,
N
.
T
o
me
h
,
a
n
d
T
.
C
h
a
r
n
o
i
s,
“
LI
P
N
a
t
W
o
j
o
o
d
N
ER
s
h
a
r
e
d
t
a
s
k
:
a
sp
a
n
-
b
a
s
e
d
a
p
p
r
o
a
c
h
f
o
r
f
l
a
t
a
n
d
n
e
s
t
e
d
A
r
a
b
i
c
n
a
me
d
e
n
t
i
t
y
r
e
c
o
g
n
i
t
i
o
n
,
”
i
n
Pro
c
e
e
d
i
n
g
s
o
f
Ara
b
i
c
N
L
P
2
0
2
3
,
S
t
r
o
u
d
s
b
u
r
g
,
P
A
,
U
S
A
:
A
sso
c
i
a
t
i
o
n
f
o
r
C
o
m
p
u
t
a
t
i
o
n
a
l
L
i
n
g
u
i
s
t
i
c
s,
2
0
2
3
,
p
p
.
7
8
9
–
7
9
6
.
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
2
0
2
3
.
a
r
a
b
i
c
n
l
p
-
1
.
8
9
.
[
1
5
]
I
.
La
o
u
i
r
i
n
e
,
H
.
El
l
e
u
c
h
,
a
n
d
F
.
B
o
u
g
a
r
e
s
,
“
ELY
A
D
A
TA
a
t
W
o
j
o
o
d
N
E
R
s
h
a
r
e
d
t
a
sk
:
d
a
t
a
a
n
d
m
o
d
e
l
-
c
e
n
t
r
i
c
a
p
p
r
o
a
c
h
e
s
f
o
r
A
r
a
b
i
c
f
l
a
t
a
n
d
n
e
s
t
e
d
N
E
R
,
”
i
n
P
ro
c
e
e
d
i
n
g
s
o
f
Ar
a
b
i
c
N
L
P
2
0
2
3
,
S
t
r
o
u
d
s
b
u
r
g
,
P
A
,
U
S
A
:
A
ss
o
c
i
a
t
i
o
n
f
o
r
C
o
m
p
u
t
a
t
i
o
n
a
l
Li
n
g
u
i
s
t
i
c
s,
2
0
2
3
,
p
p
.
7
5
9
–
7
6
4
.
d
o
i
:
1
0
.
1
8
6
5
3
/
v
1
/
2
0
2
3
.
a
r
a
b
i
c
n
l
p
-
1
.
8
4
.
[
1
6
]
W
.
A
n
t
o
u
n
,
F
.
B
a
l
y
,
a
n
d
H
.
H
a
j
j
,
“
A
r
a
B
E
R
T:
t
r
a
n
sf
o
r
m
e
r
-
b
a
se
d
m
o
d
e
l
f
o
r
A
r
a
b
i
c
l
a
n
g
u
a
g
e
u
n
d
e
r
st
a
n
d
i
n
g
,
”
i
n
P
ro
c
e
e
d
i
n
g
s
o
f
t
h
e
4
t
h
W
o
r
k
sh
o
p
o
n
O
p
e
n
-
S
o
u
rce
Ara
b
i
c
C
o
r
p
o
ra
a
n
d
Pr
o
c
e
ss
i
n
g
T
o
o
l
s,
w
i
t
h
a
S
h
a
re
d
T
a
s
k
o
n
O
f
f
e
n
s
i
v
e
L
a
n
g
u
a
g
e
D
e
t
e
c
t
i
o
n
,
2
0
2
0
.
[
1
7
]
R
.
E.
S
a
l
a
h
a
n
d
L.
Q
.
B
i
n
t
i
Z
a
k
a
r
i
a
,
“
B
u
i
l
d
i
n
g
t
h
e
c
l
a
ssi
c
a
l
a
r
a
b
i
c
n
a
me
d
e
n
t
i
t
y
r
e
c
o
g
n
i
t
i
o
n
c
o
r
p
u
s
(
C
A
N
ER
C
o
r
p
u
s)
,
”
i
n
2
0
1
8
Fo
u
rt
h
I
n
t
e
rn
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
I
n
f
o
rm
a
t
i
o
n
R
e
t
r
i
e
v
a
l
a
n
d
K
n
o
w
l
e
d
g
e
Ma
n
a
g
e
m
e
n
t
(
C
A
MP
)
,
I
EEE,
M
a
r
.
2
0
1
8
,
p
p
.
1
–
8
.
d
o
i
:
1
0
.
1
1
0
9
/
I
N
F
R
K
M
.
2
0
1
8
.
8
4
6
4
8
2
0
.
[
1
8
]
N
.
B
o
u
d
j
e
l
l
a
l
e
t
a
l
.
,
“
A
B
i
o
N
ER
:
A
B
ER
T‐
b
a
se
d
mo
d
e
l
f
o
r
A
r
a
b
i
c
b
i
o
me
d
i
c
a
l
n
a
me
d
‐
e
n
t
i
t
y
r
e
c
o
g
n
i
t
i
o
n
,
”
C
o
m
p
l
e
x
i
t
y
,
v
o
l
.
2
0
2
1
,
n
o
.
1
,
J
a
n
.
2
0
2
1
,
d
o
i
:
1
0
.
1
1
5
5
/
2
0
2
1
/
6
6
3
3
2
1
3
.
[
1
9
]
N
.
A
l
saa
r
a
n
a
n
d
M
.
A
l
r
a
b
i
a
h
,
“
A
r
a
b
i
c
n
a
me
d
e
n
t
i
t
y
r
e
c
o
g
n
i
t
i
o
n
:
a
B
E
R
T
-
B
G
R
U
a
p
p
r
o
a
c
h
,
”
C
o
m
p
u
t
e
rs,
M
a
t
e
r
i
a
l
s
&
C
o
n
t
i
n
u
a
,
v
o
l
.
6
8
,
n
o
.
1
,
p
p
.
4
7
1
–
4
8
5
,
2
0
2
1
,
d
o
i
:
1
0
.
3
2
6
0
4
/
c
m
c
.
2
0
2
1
.
0
1
6
0
5
4
.
[
2
0
]
A
.
M
e
k
k
i
,
I
.
Zr
i
b
i
,
M
.
El
l
o
u
z
e
,
a
n
d
L.
H
.
B
e
l
g
u
i
t
h
,
“
N
a
me
d
e
n
t
i
t
y
r
e
c
o
g
n
i
t
i
o
n
o
f
T
u
n
i
si
a
n
A
r
a
b
i
c
u
si
n
g
t
h
e
B
i
-
LST
M
-
C
R
F
mo
d
e
l
,
”
I
n
t
e
rn
a
t
i
o
n
a
l
J
o
u
rn
a
l
o
n
Art
i
f
i
c
i
a
l
I
n
t
e
l
l
i
g
e
n
c
e
T
o
o
l
s
,
v
o
l
.
3
3
,
n
o
.
0
2
,
M
a
r
.
2
0
2
4
,
d
o
i
:
1
0
.
1
1
4
2
/
S
0
2
1
8
2
1
3
0
2
3
5
0
0
6
2
8
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
5
0
2
-
4
7
52
I
n
d
o
n
esian
J
E
lec
E
n
g
&
C
o
m
p
Sci
,
Vo
l.
42
,
No
.
1
,
Ap
r
il
20
26
:
1
42
-
14
8
148
[
2
1
]
M
.
A
l
-
Q
u
r
i
s
h
i
a
n
d
R
.
S
o
u
i
ss
i
,
“
A
r
a
b
i
c
n
a
m
e
d
e
n
t
i
t
y
r
e
c
o
g
n
i
t
i
o
n
u
si
n
g
t
r
a
n
sf
o
r
mer
-
b
a
se
d
-
C
R
F
m
o
d
e
l
,
”
I
C
N
L
S
P
2
0
2
1
-
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
4
t
h
I
n
t
e
r
n
a
t
i
o
n
a
l
C
o
n
f
e
re
n
c
e
o
n
N
a
t
u
r
a
l
L
a
n
g
u
a
g
e
a
n
d
S
p
e
e
c
h
Pr
o
c
e
ssi
n
g
,
p
p
.
2
6
2
–
2
7
1
,
2
0
2
1
.
[
2
2
]
W
.
A
n
t
o
u
n
,
F
.
B
a
l
y
,
a
n
d
H
.
H
a
j
j
,
“
A
R
A
ELE
C
T
R
A
:
p
r
e
-
t
r
a
i
n
i
n
g
t
e
x
t
d
i
s
c
r
i
mi
n
a
t
o
r
s
f
o
r
A
r
a
b
i
c
l
a
n
g
u
a
g
e
u
n
d
e
r
s
t
a
n
d
i
n
g
,
”
Pro
c
e
e
d
i
n
g
s
o
f
t
h
e
S
i
x
t
h
Ara
b
i
c
N
a
t
u
r
a
l
L
a
n
g
u
a
g
e
Pr
o
c
e
ss
i
n
g
Wo
r
k
sh
o
p
,
p
p
.
1
9
1
–
1
9
5
,
2
0
2
1
.
[
2
3
]
R
.
S
a
l
a
h
,
M
.
M
u
k
r
e
d
,
L
.
Q
.
B
i
n
t
i
Z
a
k
a
r
i
a
,
a
n
d
F
.
A
.
M
.
A
l
-
Y
a
r
i
mi
,
“
A
ma
c
h
i
n
e
l
e
a
r
n
i
n
g
a
p
p
r
o
a
c
h
f
o
r
n
a
m
e
d
e
n
t
i
t
y
r
e
c
o
g
n
i
t
i
o
n
i
n
c
l
a
ss
i
c
a
l
A
r
a
b
i
c
n
a
t
u
r
a
l
l
a
n
g
u
a
g
e
p
r
o
c
e
ssi
n
g
,
”
K
S
I
I
T
ra
n
sa
c
t
i
o
n
s
o
n
I
n
t
e
rn
e
t
a
n
d
I
n
f
o
rm
a
t
i
o
n
S
y
s
t
e
m
s
,
v
o
l
.
1
8
,
n
o
.
1
0
,
p
p
.
2
8
9
5
–
2
9
1
9
,
O
c
t
.
2
0
2
4
,
d
o
i
:
1
0
.
3
8
3
7
/
t
i
i
s.
2
0
2
4
.
1
0
.
0
0
5
.
[
2
4
]
M
.
J
a
r
r
a
r
,
N
.
H
a
b
a
s
h
,
F
.
A
l
r
i
maw
i
,
D
.
A
k
r
a
,
a
n
d
N
.
Za
l
mo
u
t
,
“
C
u
r
r
a
s:
a
n
a
n
n
o
t
a
t
e
d
c
o
r
p
u
s
f
o
r
t
h
e
P
a
l
e
s
t
i
n
i
a
n
A
r
a
b
i
c
d
i
a
l
e
c
t
,
”
L
a
n
g
u
a
g
e
Re
s
o
u
rc
e
s
a
n
d
Ev
a
l
u
a
t
i
o
n
,
v
o
l
.
5
1
,
n
o
.
3
,
p
p
.
7
4
5
–
7
7
5
,
S
e
p
.
2
0
1
7
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
0
5
7
9
-
0
1
6
-
9
3
7
0
-
7.
[
2
5
]
K
.
E
l
H
a
f
f
,
M
.
Jarr
a
r
,
T.
H
a
mm
o
u
d
a
,
a
n
d
F
.
Za
r
a
k
e
t
,
“
C
u
r
r
a
s
+
B
a
l
a
d
i
:
t
o
w
a
r
d
s
a
l
e
v
a
n
t
i
n
e
c
o
r
p
u
s,”
Pr
o
c
e
e
d
i
n
g
s
o
f
t
h
e
T
h
i
r
t
e
e
n
t
h
L
a
n
g
u
a
g
e
R
e
s
o
u
r
c
e
s
a
n
d
Ev
a
l
u
a
t
i
o
n
C
o
n
f
e
re
n
c
e
,
p
p
.
7
6
9
–
7
7
8
,
2
0
2
2
.
B
I
O
G
RAP
H
I
E
S
O
F
AUTH
O
RS
S
o
m
i
a
K
h
e
d
im
i
e
a
rn
e
d
h
e
r
m
a
ste
r’s
d
e
g
re
e
i
n
c
o
m
p
u
ter
sc
ien
c
e
fro
m
t
h
e
Un
iv
e
rsity
o
f
Na
a
m
a
in
2
0
1
8
.
S
i
n
c
e
2
0
2
3
,
s
h
e
h
a
s
b
e
e
n
p
u
rs
u
in
g
a
P
h
.
D.
i
n
c
o
m
p
u
ter
sc
ien
c
e
a
t
th
e
sa
m
e
in
stit
u
ti
o
n
.
He
r
r
e
se
a
rc
h
in
tere
sts
fo
c
u
s
o
n
a
rti
ficia
l
in
telli
g
e
n
c
e
a
n
d
it
s
a
p
p
li
c
a
ti
o
n
s,
wit
h
p
a
rt
icu
l
a
r
e
m
p
h
a
sis
o
n
n
a
tu
ra
l
lan
g
u
a
g
e
p
ro
c
e
ss
in
g
.
S
h
e
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
k
h
e
d
imi@c
u
n
iv
-
n
a
a
m
a
.
d
z
.
Abd
e
lg
h
a
n
i
B
o
u
z
ia
n
e
is
a
n
a
ss
o
c
iate
p
ro
fe
ss
o
r
o
f
c
o
m
p
u
t
e
r
sc
ien
c
e
a
t
th
e
Un
iv
e
rsity
o
f
Na
a
m
a
.
His
a
c
a
d
e
m
ic
e
x
p
e
rti
se
sp
a
n
s
a
rti
f
icia
l
in
t
e
ll
ig
e
n
c
e
,
m
a
c
h
in
e
lea
rn
in
g
,
d
e
e
p
lea
rn
i
n
g
,
a
n
d
n
a
tu
ra
l
lan
g
u
a
g
e
p
r
o
c
e
ss
in
g
.
He
o
b
tain
e
d
h
is
P
h
.
D.
i
n
c
o
m
p
u
ter
sc
ien
c
e
fro
m
th
e
U
n
iv
e
rsit
y
o
f
S
i
d
i
Be
l
Ab
b
è
s,
Alg
e
ria,
in
2
0
1
9
.
S
in
c
e
2
0
1
2
,
h
e
h
a
s
b
e
e
n
a
c
ti
v
e
ly
e
n
g
a
g
e
d
in
a
d
v
a
n
c
in
g
a
n
d
i
n
teg
r
a
ti
n
g
a
rti
ficia
l
i
n
telli
g
e
n
c
e
tec
h
n
o
lo
g
ies
wit
h
in
b
o
th
re
se
a
rc
h
a
n
d
tea
c
h
in
g
c
o
n
tex
ts.
At
th
e
Un
iv
e
rsity
o
f
Na
a
m
a
,
h
e
h
o
l
d
s
th
e
p
o
siti
o
n
s
o
f
h
e
a
d
o
f
fiel
d
i
n
th
e
c
o
m
p
u
ter
sc
ien
c
e
d
e
p
a
rtme
n
t
a
n
d
c
h
ief
o
f
d
o
c
to
ra
l
trai
n
in
g
i
n
a
rti
ficia
l
in
telli
g
e
n
c
e
.
Hi
s
re
se
a
rc
h
in
tere
sts
a
r
e
p
rima
ril
y
c
e
n
tere
d
o
n
a
rti
ficia
l
in
telli
g
e
n
c
e
a
n
d
n
a
tu
ra
l
lan
g
u
a
g
e
p
ro
c
e
ss
in
g
.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
b
o
u
z
ian
e
@c
u
n
i
v
-
n
a
a
m
a
.
d
z
.
Evaluation Warning : The document was created with Spire.PDF for Python.